HTI convened 30 experts in public policy, emerging technologies, civic engagement, and information studies to discuss how AI is impacting democratic resilience in Australia
How artificial intelligence is impacting democratic resilience
Artificial Intelligence (AI) is transforming how Australians communicate, form social relationships, engage with public services, and receive information. Cumulatively, these shifts can potentially disrupt three key pillars of Australian democracy: trust in public institutions, access to credible information and the inclusiveness of society.
To better understand the dynamics of these shifts, HTI convened 30 experts in public policy, emerging technologies, civic engagement, and information studies to discuss how AI is impacting democratic resilience in Australia. The session took place on Monday 16 October 2023 under Chatham House Rule in collaboration with the Department of Home Affairs.
During the discussion, participants observed that it is easier to think of ways that AI is undermining democracy than it is to provide examples of how AI is improving democratic processes and institutions. However, participants observed that this doesn’t have to be the case.
Roundtable participants discussed the threats and opportunities posed by AI to Australian democratic processes. For each of these areas, participants identified priority areas for research and policy to protect and promote Australia’s democratic resilience in the age of AI.
Key insights
Australian democracy faces new challenges thanks to AI systems that are undermining the credibility of public information and creating social division.
Access to news, public information and civic debate largely now occurs online, with the flow of information increasingly fractured by algorithmic automation and micro-targeting. The rapid rise of mis- and disinformation online, amplified by generative AI tools that can create synthetic news and deepfake images, videos and soundbites, challenges citizen access to credible information and the ability to engage constructively in democratic processes and civic debates.
Participants touched on the role that digital fingerprinting, AI-literacy and AI-based learning systems could play in supporting both organisations and citizens to validate public information. Participants also noted the possibility for AI systems to be designed to facilitate new forms of civic participation and engagement, including the synthesis of information to help frame common challenges and help increase access to services.
AI systems have the potential to increase trust in public institutions by making government services more reliable and higher quality – but this will require new accountability mechanisms and governance processes.
Both poor AI system performance and the misuse of AI may further lower public trust in Australian institutions. This is a particular risk in situations where the government fails to appropriately govern AI systems that engage directly with citizens or contribute to administrative decisions.
Participants saw a range of opportunities to strengthen public accountability mechanisms to ensure the use of AI is accurate, fair and reliable. These included AI Assurance mechanisms, transparency and contestability requirements, enhanced privacy protections, and co-designing systems with community
A concerted effort is required to ensure that the design of AI systems supports social inclusion and facilitates democratic participation and engagement.
Participants noted that an increasing range of existing applications of AI, including allocating work in the gig economy, determining creditworthiness for loans and multiple forms of digital surveillance, risk further entrenching and amplifying existing social inequalities.
Participants noted the importance of putting people and their needs at the centre of AI system purpose and design, for example by ensuring that systematically managing AI system harms people is a key priority.
Ensuring that AI systems uphold human values and rights is at the heart of HTI’s work on emerging technologies. For further information on related work, see HTI’s projects on facial recognition technology, digital identity and digital government, AI corporate governance, and the future of AI regulation.